AI Integration

Experience

FinTech

• Building an agentic AI serving platform centered around observability and memory management
• Designing LLM tracing and observatory pipelines connected to knowledge management systems
• Developing multi-agent orchestration for financial automation workflows
• Evaluating and automating RAG pipelines and agent evaluation frameworks

Cloud-Native AI

• Fine-tuned open-source LLMs for domain-specific reasoning
• Designed distributed LLM training pipelines on multi-cloud environments
• Automated ML lifecycle on Kubernetes with custom operators

Automotive

• Built RAG applications for automotive consultation and technical documentation
• Developed custom parsing and preprocessing modules for diverse document formats
• Built data ingestion pipelines for multi-source automotive datasets
• Integrated corporate data across multiple European countries under GDPR constraints

Open Source

• Graphiti — Knowledge graph construction for AI agents
• Serena — Semantic code retrieval and editing toolkit (MCP server)
• Aristotle — Blockchain-based artwork platform (own project)

Research Interests

Fine-tuning for domain adaptation · Human-in-the-loop agent design · Agent memory architectures · LLM observability and tracing Medium: @liebertar

Activity